Summary of work: Studies of gender differences and individual variability in age-related phenomena are being carried out to: 1) determine the "normal" range of variability in human aging, 2) identify potential sources of variability which may be responsive to intervention, and 3) determine if there are subgroups of individuals who are more susceptible or resistant to various aspects of aging. The research combines the use of sophisticated statistical methodologies and the unique time depth and multidisciplinary breadth of the existing BLSA data base to examine issues related to the concepts of "normal" and "successful" aging, as well as to increase the power of traditional research designs. The statistical methods used include longitudinal regression models, time dependent proportional hazards analysis, and finite mixture models. Major findings include that screened samples of BLSA men and women change in hearing threshold relative to the baseline audiogram of an average of 10 dB or more at 2, 3 and 4 kHz in either ear increase with age and follow-up time in men, but little change occurs in women at all ages. The resulting percentile curves represent norms for changes in hearing level in the absence of any known otologic disease. These percentile curves provide a reference for detecting when a person deviates from a normal pattern of change, thus helping in diagnosing problems with hearing or in monitoring hearing in occupational settings. These findings represent significant contributions to the theoretical and methodological development of biomedical risk factor studies, as well as to an increased understanding of the dynamics of the aging process. Research is underway to develop more refined methods of studying variability in aging in order to develop theoretically and methodologically sound approaches to risk factor analysis which account for changes in an individual's covariates over time and the possibility that individuals differ in susceptibility or resistance to aging processes.